|
on Discrete Choice Models |
| By: | Klaus Moeltner (Department of Resource Economics, University of Nevada, Reno); James J. Murphy (Department of Resource Economics & Center for Public Policy and Administration, University of Massachusetts-Amherst); John K. Stranlund (Department of Resource Economics, University of Massachusetts-Amherst); Maria Alejandra Velez (Center for Research on Environmental Decisions, Columbia University) |
| Abstract: | Economists have to date not been very discriminating in selecting parametric models to analyze experimental data. In this study we propose two parametric estimators tailored to accommodate both the bounded integer outcomes and the latent subject heterogeneity typically observed for Social Dilemma games. The two estimators are the Hierarchical Ordered Probit and the Hierarchical Doubly-Truncated Poisson. We illustrate how both specifications can be implemented in a Bayesian estimation framework, which circumvents estimation hurdles and allows for maximum flexibility in model comparison and model selection. We apply this framework to data from a Common Pool Resource game implemented in rural communities. We find that both estimators capture the essential effects and patterns present in the data. As expected, the truncated count data framework exhibits higher efficiency in estimation and prediction. |
| Keywords: | Social Dilemma Games; Hierarchical Bayesian Modeling; Ordered Probit; Truncated Poisson; Common Property Resource |
| JEL: | C11 C24 C93 Q22 |
| Date: | 2007–12 |
| URL: | http://d.repec.org/n?u=RePEc:unr:wpaper:07-013&r=dcm |
| By: | Ardic, Oya Pinar; Yuzereroglu, Uygar |
| Abstract: | This paper uses a multinomial probit model to analyze individuals' choice of banks based on the types of banking services they use, their own characteristics, and their own perceptions about important factors in banking. Previous studies on this topic, which are limited in number, concentrate on the U.S. where financial markets are deep. This analysis uses a unique individual level data set from a nation-wide survey implemented after the 2001 crisis in Turkey, of which one major component was bank failures. Hence, it provides the first set of evidence on the topic in an emerging market context. The study groups banks into three categories: public banks, large private banks and small private banks, among which the latter is perceived to be the potentially risky group. Investigating individuals' choice among these three types, the paper uncovers that while individuals tend to prefer small private banks on the basis of high interest rates, they tend to avoid them on the basis of trust. However, higher branch density and closeness negatively affect the choice of small private banks. Additionally, individual's choice of public banks as opposed to large private banks seems to have been positively influenced mostly by being older, being retired, receiving salary/pension, and valuing special services for farmers and craftsmen while it seems to have been negatively influenced by the use of certain services, valuing friendliness of the staff, and living in more developed regions where there is variety in terms of the financial institutions. |
| Keywords: | Multinomial probit; banking sector; bank choice; household survey; Turkey |
| JEL: | D12 C25 G21 |
| Date: | 2007–11–08 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:6096&r=dcm |
| By: | Constantina Kottaridi; Fragkiskos Filippaios |
| Abstract: | Foreign Direct Investment (FDI) and its agents, i.e. Multinational Corporations (MNCs), are understood to play a major role in the economic development of nations through their impact on trade and their ability to generate jobs and to produce new knowledge through technological and managerial advances (UNCTC, 2003). This article maps economic activities of MNCs in United Kingdom and presents an empirical formulation of investors’ decision-making. A McFadden’s conditional logit model was incorporated to test for the model’s predictions, based on location decisions of 6348 plants in UK’s counties for 2004. Estimation results suggest that firms’ choices can be modelled in terms of economic factors prevailing locally. |
| Keywords: | UK Regions, subsidiaries, environmental determinism, location choice |
| Date: | 2007 |
| URL: | http://d.repec.org/n?u=RePEc:uop:wpaper:010&r=dcm |
| By: | Mohammad Alauddin; Clem Tisdell (School of Economics, The University of Queensland) |
| Abstract: | This paper explores the factors that affect students’ evaluations of economics instructions using a sample of over 2400 completed questionnaires at a large Australian university. Ordered probit analysis is used to determine the changes in the predicted probability of teaching evaluation (TEVAL) scores with variations, amongst other things, in students’ perceptions of the quality of presentation; explanation and organization of lecture material; and helping students improve their learning skills. Analyses of the comparative importance of the relationships both for undergraduate and postgraduate courses reveal significant differences across levels of the undergraduate program but little differences in students’ responses in higher level undergraduate and postgraduate instructions. One disturbing finding is that a key variable, namely emphasis on thinking rather than memorizing (THINKMEM) has little or no substantive impact on TEVAL. Thus the implication is that high TEVALs can be achieved at the cost of some critically important factors in teaching and learning. Consequently, those using just TEVAL score to evaluate teaching need to look closely at other factors of critical importance. |
| Date: | 2007 |
| URL: | http://d.repec.org/n?u=RePEc:qld:uq2004:353&r=dcm |